Abstract
Abstract. Temperature exerts strong controls on the incidence and severity of fire. All else equal, warming is expected to increase fire-related carbon emissions, and thereby atmospheric CO2. But the magnitude of this feedback is very poorly known. We use a single-box model of the land biosphere to quantify this positive feedback from satellite-based estimates of biomass burning emissions for 2000–2014 CE and from sedimentary charcoal records for the millennium before the industrial period. We derive an estimate of the centennial-scale feedback strength of 6.5 ± 3.4 ppm CO2 per degree of land temperature increase, based on the satellite data. However, this estimate is poorly constrained, and is largely driven by the well-documented dependence of tropical deforestation and peat fires (primarily anthropogenic) on climate variability patterns linked to the El Niño–Southern Oscillation. Palaeo-data from pre-industrial times provide the opportunity to assess the fire-related climate–carbon-cycle feedback over a longer period, with less pervasive human impacts. Past biomass burning can be quantified based on variations in either the concentration and isotopic composition of methane in ice cores (with assumptions about the isotopic signatures of different methane sources) or the abundances of charcoal preserved in sediments, which reflect landscape-scale changes in burnt biomass. These two data sources are shown here to be coherent with one another. The more numerous data from sedimentary charcoal, expressed as normalized anomalies (fractional deviations from the long-term mean), are then used – together with an estimate of mean biomass burning derived from methane isotope data – to infer a feedback strength of 5.6 ± 3.2 ppm CO2 per degree of land temperature and (for a climate sensitivity of 2.8 K) a gain of 0.09 ± 0.05. This finding indicates that the positive carbon cycle feedback from increased fire provides a substantial contribution to the overall climate–carbon-cycle feedback on centennial timescales. Although the feedback estimates from palaeo- and satellite-era data are in agreement, this is likely fortuitous because of the pervasive influence of human activities on fire regimes during recent decades.
Highlights
Fire is a natural, recurring event in most terrestrial ecosystems
Climateinduced inter-annual variability in biomass burning, variability associated with the El Niño–Southern Oscillation (ENSO), is an important component of the inter-annual variability of the atmospheric CO2 growth rate
Changes in biomass burning need to be taken into account in estimating the “climate–carbon-cycle feedback”, i.e. the longer-term positive feedback by which global warming leads to a reduction in land carbon storage, a consequent reduction in the net uptake of CO2 so that more CO2 remains in the atmosphere, and an amplification of the initial warming (Arora et al, 2013; Cox et al, 2013; Wenzel et al, 2014)
Summary
Fire is a natural, recurring event in most terrestrial ecosystems. About 4 % of the global land area is burnt every year (Giglio et al, 2013), resulting in global CO2 emissions of around 2 PgC per year (van der Werf et al, 2010), substantial contributions to the budgets of other direct or indirect greenhouse gases (including CH4, CO, N2O, and ozone precursors), and further contributions to the atmospheric aerosol loading (black carbon, organic compounds). Sensed observations of biomass burning offer a uniquely detailed global perspective on fire regimes They cover only a limited period, and our ability to use these records to derive an empirical estimate of the biomass burning contribution to the carbon cycle feedback is further compromised by the complexity of the controls on fire. CH4 is released during the smouldering phase of fires, roughly in proportion to total CO2 emission (Andreae and Merlet, 2001) This process is a relatively minor contributor to total atmospheric CH4, it disproportionately influences the 13C content of CH4 because pyrogenic CH4 carries the isotopic signature of photosynthesis. We use a single-box model of the land biosphere to derive an estimate of the contemporary biomass burning contribution to the climate–carbon-cycle feedback using remote-sensingbased estimates of biomass burning carbon emissions for the interval 2000–2014 CE. We exploit a good correlation of normalized anomalies of global charcoal abundance with global land temperatures during the last millennium to derive an alternative estimate of the strength of the climate–carbon-cycle feedback
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
